W3C home > Mailing lists > Public > public-lod@w3.org > May 2010

Re: [Virtuoso-users] [Fwd: How to serve hash URIs from a (Virtuoso) SPARQL endpoint?]

From: Carl Blakeley <cblakeley@openlinksw.co.uk>
Date: Wed, 19 May 2010 17:49:44 +0100
Message-ID: <4BF416A8.2090500@openlinksw.co.uk>
To: Christoph LANGE <ch.lange@jacobs-university.de>
CC: Hugh Williams <hwilliams@openlinksw.com>, Virtuoso Users List <virtuoso-users@lists.sourceforge.net>, "public-lod@w3.org" <public-lod@w3.org>, "nikita.zhiltsov@gmail.com" <nikita.zhiltsov@gmail.com>, "Zholudev, Vyacheslav V." <v.zholudev@jacobs-university.de>
Christoph LANGE wrote:
> Hi Hugh,
>
> [@Nathan, thanks for forwarding it to the Virtuoso list; I subscribed there
> now, but I thought the question might also be of a more general interest
> w.r.t.  deploying hash URIs.]
>
> 2010-05-19 14:51 Hugh Williams <hwilliams@openlinksw.com>:
>   
>> The following Virtuoso Linked Data Deployment Guide details how hash URIs can be handled by the server using transparent content negotiation:
>>
>> 	http://www.openlinksw.com/virtuoso/Whitepapers/html/vdld_html/VirtDeployingLinkedDataGuide.html
>>
>> Which would also apply to the data you are attempting to publish ...
>>     
>
> Thanks, I had looked there before, but got the impression that that guide only
> deals with the very special case of the "this" pseudo fragment ID, i.e. a
> workaround of introducing hash URIs to facilitate content negotiation.  I got
> that impression because the guide talks about http://.../ALFKI#this, where
> "ALFKI" is the entity of interest.  Please let me know if I got something
> wrong.
>
> In our situation, we have many entities of interest, with the following URIs:
>
> http://.../document (without fragment)
> http://.../document#fragment1
> http://.../document#fragment2
> ...
>
> and when a client requests RDF/XML from http://.../document, the client should
> get a document that contains all triples for http://.../document,
> http://.../document#fragment1, http://.../document#fragment2, etc.
>
> (Note that we were not free to choose this URI format; it was given before we
> went "linked data".)
>
> Cheers, and thanks,
>
> Christoph
>
>   
Christoph,

Whatever the SPARQL query required to return all triples of the form 
http://.../document[#fragment] in response to a request for 
http://.../document,

- be it Mischa's suggestion (if you're able to load all triples with 
subject http://.../document[#fragment] into graph http://.../data/document):

construct {?s ?p ?o} where
{graph <http://our.server/data/document> {?s ?p ?o}}

- or the original query (if you're not):

construct { ?s ?p ?o }
where { ?s ?p ?o .
    filter(
        ?s = "http://our.server/data/document"
            ||
        regex(?s,"^http://our.server/data/document#"))
}

you should be able to hide either behind a Virtuoso rewrite rule for 
http://.../document.

Regards,
Carl Blakeley
OpenLink Software
Received on Wednesday, 19 May 2010 20:35:15 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:29:48 UTC